Document Type

Working Paper

Publication Date



This paper describes two initiatives that have used experimental studies to guide the development of community-based health and family planning programs. In Bangladesh and Ghana, factorial experiments were implemented in stages. An exploratory phase developed a service system for community-based health care; an experimental phase assessed the demographic impact of the system; a replication phase examined the transferability of the experimental program to a non-research setting; and a scaling-up phase facilitated the extension of the new system to the national health care program. All stages were guided by research, with questions, mechanisms, and outcomes shifting as the process developed. Large-scale systems development was achieved in both Bangladesh and Ghana, not because the scaling-up programs were alike, but because similar research approaches informed their strategies, allowing them to adapt to contrasting societal and institutional contexts. Success in Bangladesh and Ghana suggests ways in which evidence-based system development can overcome resource and organizational constraints and foster transitions from limited, passive clinical services to active programs for providing accessible community-based care.